Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Journal of Geographic Issues, Vol. 2 No. 1, 2014
58
EMPIRICAL ESTIMATION OF GROUNDWATER RECHARGE IN PARTS OF
THE BASEMENT COMPLEX OF NORTHERN NIGERIA
Olubunmi Adegun
Department of Geography, University of Lagos, Lagos, Nigeria
Correspondence:[email protected]
Abstract Knowledge of recharge estimates is essential to the sustainable development and
management of groundwater resources. In this study two empirical methods were employed
in estimating of groundwater recharge in Gusau and Bauchi in semi-arid Northern Nigeria.
The recharge estimates were based on the methods developed by Sinha and Sharma (1988)
and Turc (1954).Variability of annual rainfall and recharge estimates were determined using
the coefficient of variation, while the equality of variance of recharge estimates obtained
from the empirical methods were determined using Levnne’s Test. Simple Linear Regression
was used to predict the mean groundwater recharge for the study period. Results of
groundwater recharge for Bauchi based on the method developed by Sinha and Sharma gave
a maximum and minimum estimate of 262.64mm and 158.50mm, while the highest and
lowest estimate for Gusau was 252.49mm and 142.29mm. Highest and lowest annual average
recharge estimate based on Turc’s method was, 687.49mm and 170.13mm for Bauchi, while
the maximum and minimum estimate for Gusau was 578.76mm and 69.52mm. The results of
the coefficient of variation shows a high level of variability based on Turc’s method, and a
moderate degree of variability based on Sinha and Sharma’s method. The results of Levene’s
Test indicated that variance of recharge estimates obtained from both methods differ
significantly from each other. Regression model for the two locations revealed that rainfall
generally accounts for more than 90 percent of groundwater recharge variation. The study
recommended that scientific and periodic assessments of groundwater resources potential
should be embarked upon.
Keywords: Recharge; Groundwater; Semi-Arid; Basement Complex; Rainfall.
INTRODUCTION Groundwater recharge is a process by which water enters into the phreatic or saturated zone.
It is the downward flow or percolation of water to the water table, adding to water in the
groundwater reservoir (Sharp, 2007; de Vries and Simmers, 2002). The highly variable
nature of groundwater recharge and its pivotal role in the sustainable development and
management of groundwater resources therefore makes the quantification of natural recharge
very important. In arid and semi-arid regions of the world where rates of natural groundwater
recharge are generally low in comparison with average annual rainfall, quantification of the
rate of natural groundwater recharge is necessary for the sustainable use and management of
groundwater.
In the semi-arid areas of Northern Nigeria where groundwater is being increasingly utilized,
the estimation of groundwater available for use is very important in order to prevent some of
the negative environment consequences associated with over-abstraction and mining of
groundwater. The increased reliance on groundwater in many parts of Northern Nigeria is
attributed to a number of reasons. These include the highly seasonal nature of rainfall, low
rainfall amounts and water balance deficit, increasing population, inadequate pipe-borne
water supply for domestic and municipal uses, and over- abstraction of groundwater for
Journal of Geographic Issues, Vol. 2 No. 1, 2014
59
agriculture and other uses. Furthermore, the downward trend of river discharge occasioned by
the sahelian drought of the 1970s and 1990s also contributed to increased dependence on
groundwater within the region as groundwater became the only reliable source of water
during those periods.
The combined effects of climate variability, the generally low yields of crystalline basement
rocks and heavy dependence on groundwater has often resulted in the drying up of shallow
wells and boreholes, and decline in groundwater levels, which eventually culminates in
higher pumping heads and greater costs of pumping. This situation therefore raises a lot of
concerns about the sustainable development, use and management of groundwater in the
region. As a starting point in the sustainable management of this vital resource, working
estimates of the rates of groundwater recharge is required.
Numerous techniques and models have been developed over the years, to estimate natural
groundwater recharge. These include the empirical method, groundwater resource estimation
method, groundwater balance approach and soil moisture based methods (Kumar,
2000).Other methods as identified by Kumar & Seethapathi (2002), include groundwater
balance method, soil water balance method, zero flux plane method, one-dimensional soil
water flow model, inverse modeling technique and isotope and solute profile techniques.
A number groundwater recharge studies in which empirical methods were used exists in
literature. For example, Oke et al., (2013) compared the estimated annual groundwater
recharge for the Ogun-Oshun River Basin using three empirical methods, namely the Krishna
Rao formula, the Kumar and Seethapathi formula and the modified Chaturvedi formula. The
results of the study showed that the annual recharge estimate based the first two formulae
were similar while the estimates based on the modified Chaturvedi formula differed
significantly.
Utilising the Amristsar formula in combination with remote sensing techniques, Rawat et al.,
(2012) estimated groundwater recharge at Shankergarh block in Allahabad, India. The result
of the study revealed a close correlation between the empirical method employed and the
estimate obtained through remote sensing. Adeleke et al., (2015) used the modified
Chaturvedi formula to estimate groundwater recharge for Odeda Local Government Area of
Ogun State between 1983 and 2014. The result of the study showed that the rate of recharge
was low with 11 percent of the rainfall received percolating into the underlying aquifers.
Furthermore, a number of studies exist on groundwater recharge in Northern Nigeria in which
different estimation methods were used. These include the adoption of the water budget
method in estimating groundwater recharge in the Hadejia-Nguru wetlands (Goes, 1999), the
use of the chloride mass balance method in estimating groundwater recharge in Northeast
Nigeria (Edmunds, et al, 2002), use of the unsaturated zone chloride profile technique in
estimating groundwater recharge in Northeast Nigeria (Goni, 2001) and the utilization of a
combination of empirical, chloride mass balance, climatic-hydrological balance, stream flow
hydrographs and water table recharge methods, in estimating groundwater recharge in the
parts of Sokoto Basin (Adelana, et al, 2006).
In recognition of the high variability of this natural flux and the important role that
groundwater availability plays in the economic sustenance of Northern Nigeria, this study
aims to estimate natural groundwater recharge at Gusau (Northwest Nigeria) and Bauchi
Journal of Geographic Issues, Vol. 2 No. 1, 2014
60
(Northeast Nigeria) located in the basement complex of Northern Nigeria using empirical
methods. Secondly, the study aims to determine the level of variability of annual rainfall and
the recharge estimates. Thirdly, the study aims to determine the existence or otherwise of
equality or homogeneity in the variance of the recharge estimates derived from the empirical
methods. Lastly the study aims to assess the extent to which rainfall accounts for variation in
recharge over the study period as well as predict the mean groundwater recharge for the study
period using the simple linear regression method.
MATERIALS AND METHODS
The study locations are Gusau and Bauchi. Gusau (Fig.1) is located between latitudes
12013′N and 12
018′N and between longitudes 6
029′E and 6
045′E (Dalhatu & Garba, 2012).
Gusau Local Government Area in which Gusau Town is located occupies an area of 3, 364
Km2 (Ladan, et al., 2012). Gusau is located in the Sudan Savanna ecological zone of Nigeria.
Climate-wise the dry season lasts from November to April, while the rainy season is from
May to October. The rainfall pattern is characterized by a single peak, with an annual mean
of about 875mm, while mean annual temperature is around 300C (Dalhatu & Garba, 2012).
The hydrographic features of Gusau area include River Sokoto and some of its tributaries, as
well as some of the westward and southward flowing tributaries of Rivers Rima and Niger
(Shear & Partners as cited in Dalhatu & Garba, 2012). Relief-wise Gusau is located at an
altitude of 300m above sea level (Dalhatu & Garba, 2012). The topography of the area is
undulating and slopes gently into the river valley (Swindel as cited in Dalhatu & Garba,
2012).
Bauchi town is located between latitudes 9000′N and 9
030′N and between longitudes 10
025′E
and 11020′E. It occupies a total land area of 3, 604 hectares (36.04Km
2). Bauchi is located in
the Sudan Savanna zone of Nigeria. The landscape is relatively flat, with a range of disjointed
hills at the northern part of the town (Usman & Mohammed, 2012).
Rainfall in Bauchi is strongly seasonal, starting mostly in the first half of May and ending in
September. Rain days vary between 100 and 160, while annual rainfall is about 1,000mm.
Highest temperatures are recorded in April and May prior to the onset of the rainy season
when mean monthly is about 29.70C while mean monthly temperature drops to 22.2
0C in the
month of December (Hill & Wall, 1978; Hazell, 1985).
Journal of Geographic Issues, Vol. 2 No. 1, 2014
61
Bauchi and Gusau are underlain by the basement complex rocks of Northern Nigeria. The
basement complex rocks are mostly made up of gneisses, migmatites and granites, with most
of the area underlain by this rock characterized by a thin, discontinuous mantle of weathered
rock (Du Preez & Barber, 1965).
Basement complex rocks are generally poor aquifers. The occurrence of groundwater
depends on the thickness and extent of the decomposed mantle or the presence of joints and
fractures in the fresh rock, with groundwater existing either in the weathered mantle or in the
joints and fracture system in the unweathered rocks.
Study on the groundwater resources development of Bauchi State (Hazell, 1985) showed that
the crystalline basement complex rocks have variable permeability, low water storage
capacity, with the depth to water table in the weathered zone ranging from 3 to 18m, while
the depth ranges between 25 and 40m where there are fractures and joints. According to
Mustafa and Adamu as cited in Dike et al (2008), hydraulic conductivity of basement
complex aquifers in Bauchi Town ranges between 0.09 and 0.46m/day, while transmissivity
is less than 2m2/day. Average borehole depth is about 32m, with a yield of about 0.8m/s
recorded in some parts of the metropolis. With regards to groundwater occurrence at Gusau,
boreholes drilled into the pre-cretaceous crystalline rocks 23 miles from Gusau showed that
the thin weathered veins made up of quartz fragments, sand and clay fillings were found to
contain most of the exploitable water at depths ranging between 94 and 135 ft.
Fig. 1: Locations of Bauchi and Gusau
Journal of Geographic Issues, Vol. 2 No. 1, 2014
62
Data used for this study are rainfall and temperature data for the period 2000 to 2010. The
data were obtained from the archives of the Nigerian Meteorological Agency (NIMET). Both
datasets served inputs into the groundwater recharge equations. Two empirical methods were
utilised in the estimation of groundwater recharge for this study. The first method was
developed by Sinha and Sharma (1988). The method utilises rainfall as the only variable and
it is applicable to semi-arid environments with rainfall greater than 380mm/year (Adelana et
al., 2006). The formula as utilised in the works of Adelana et al (2006) is expressed as:
r = 50.8 (p/25.4-15)0.4
………………………………………….(1)
Where
r = Recharge
p = Precipitation
The second empirical method is based on Turc’s (1954) empirical method for estimating
recharge. The formula is expressed as:
r = P (1-[0.9+P2/L
2]
-0.5) ………………………………..(2)
Where
r = Annual average recharge (mm/yr)
p = Annual precipitation (mm/yr)
T = Mean annual temperature (0C)
L= 300+25T+0.05T2 …………………………………………...(3)
The two methods were adopted because empirical methods have been found to be useful in
areas with inadequate hydrogeological data. Secondly, both methods were developed to
estimate groundwater recharge in similar climatic and hydrogeological environments. Both
methods were adopted as used in Adelana et al (2006), due to the similarities in
environmental conditions, as Gusau and Bauchi are located in the same semi-arid region in
which the Sokoto basin is located.
In order to determine the level of variability of annual rainfall and the estimated recharge
over the 11-year study period, the coefficient of variability was used. This is expressed as:
CV (%) �
�*100 ………………………………………….. (4)
where
σ = Standard deviation of the data series
X = Mean of the data series.
Levene’s test was used to test the equality of variance between the recharge estimates derived
from the two empirical methods. This is expressed as:
� = (� = (� − �) n�(Z��Z)�
���2
(g − 1) ∑�����
�
���(Z��
− Z�)2 ………………………………… (5)
Where
��� = |��� − ��| ………………………………………………………………… (6)
Zk=�
��∑ �� ��
��� ………………………………………………………………… (7)
� =�
! ∑ Zki$�
%���
��� …………………………………………………………… (8)
Yk=�
��∑ �� ��
��� ………………………………………………………………… (9)
Simple linear equation was used to predict the mean groundwater recharge for the study
period. The equation of the simple linear regression is expressed as:
Journal of Geographic Issues, Vol. 2 No. 1, 2014
63
Ƴ = (b0 + b1X1 ) + ɛi ............................................................................... (10)
where
Ƴ = Dependent variable predicted by regression model
b0 = Intercept
b1 = ith coefficient of Xi
Xi= ith independent variable from total set of q variables
ɛi = Random errors
RESULTS AND DISCUSSION Recharge Estimates
The annual rainfall amount and the results of recharge estimates based on Sinha and Sharma
(1988) and Turc (1954) for Bauchi and Gusau is as presented in Table 1.
Table 1: Annual Rainfall Amount and Recharge Estimates for Bauchi and Gusau
Bauchi Gusau
Year Annual
Rainfall
(mm)
Recharge Estimates
(mm)
Annual
Rainfall
(mm)
Recharge Estimates
(mm)
Sinha
and Sharma
( Method 1)
Turc
(Method 2)
Sinha
and Sharma
( Method 1)
Turc
(Method 2)
2000 1058.9 188.98 320.95 862.1 208.28 194.58
2001 1300.8 213.36 501.46 732.8 145.29 125.24
2002 819.1 158.50 170.13 889.4 210.82 210.70
2003 989.5 180.85 263.40 1416.1 252.49 578.76
2004 865.5 165.1 188.07 755.4 197.10 129.78
2005 1034.5 186.44 292.97 920.3 172.21 220.23
2006 1017.9 184.40 281.96 961.8 217.42 236.12
2007 1136.9 197.61 376.54 615.8 181.86 69.52
2008 1133.1 197.10 361.23 954.0 216.41 271.70
2009 1531.3 261.62 674.23 1006.0 221.49 284.80
2010 1547.0 262.64 687.49 1035.9 224.03 283.11
For Bauchi the results of the natural groundwater recharge estimates based on method 1
(Sinha & Sharma, 1988) shows that the periods of highest and lowest recharge coincided with
the period of highest and lowest annual rainfall amount. In 2010 when the highest annual
rainfall amount of the series (1547mm) was recorded, the highest recharge estimate of
262.64mm was also computed. For the year 2002 when the lowest annual rainfall amount
(819.1mm) of the series was recorded, the lowest recharge estimate of 158.50mm was also
computed.
The results of the groundwater recharge estimation based on the second method (Turc, 1954)
also shows that periods of highest and lowest annual average recharge estimates coincided
with periods of highest and lowest annual rainfall. Year 2010 in which the highest annual
rainfall of the series was recorded coincided with the year with the highest annual average
groundwater recharge estimate of 687.49mm. Similarly, year 2002 when the lowest annual
rainfall of the series was recorded coincided with the period of the lowest annual average
recharge estimate of 170.13mm.
Journal of Geographic Issues, Vol. 2 No. 1, 2014
Furthermore, most of the estimated annual recharge based on Sinha and Sharma method
(Fig.2) and Turc method (Fig.3) were below the 11
respectively.
Fig.2: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Sinha and
Sharma Method
Fig.3: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Turc Method
For Gusau, results of the recharge estimates
highest annual rainfall amount of the series (rainfall amount of 1416.1mm for 2003)
coincided with the period of the highest annual recharge estimate of 252.49mm. On the other
hand however, the period with the lo
2007) did not coincide with period of the lowest annual recharge estimate of 142.29mm
which was computed for 2001. The year 2001 for which the lowest annual recharge estimate
was computed coincided with the period with the second lowest annual rainfall of the series.
High potential evapotranspiration may a contributive factor to the period of lowest annual
rainfall not coinciding with the period for which lowest groundwater recharge was computed.
For the second recharge method, periods of highest and lowest annual rainfall coincided with
periods of highest and lowest annual average recharge estimates. Year 2003 in which the
highest rainfall amount of the series was recorded coincided with the period of
annual average recharge estimate of 578.76mm. The same scenario was established for the
period of lowest annual rainfall and lowest annual average recharge estimate. The period with
of Geographic Issues, Vol. 2 No. 1, 2014
64
Furthermore, most of the estimated annual recharge based on Sinha and Sharma method
(Fig.2) and Turc method (Fig.3) were below the 11-year mean of 199.7 mm and 37
Fig.2: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Sinha and
Fig.3: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Turc Method
For Gusau, results of the recharge estimates based on method 1 show that the period with
highest annual rainfall amount of the series (rainfall amount of 1416.1mm for 2003)
coincided with the period of the highest annual recharge estimate of 252.49mm. On the other
hand however, the period with the lowest annual rainfall of the series (615.8mm recorded in
2007) did not coincide with period of the lowest annual recharge estimate of 142.29mm
which was computed for 2001. The year 2001 for which the lowest annual recharge estimate
th the period with the second lowest annual rainfall of the series.
High potential evapotranspiration may a contributive factor to the period of lowest annual
rainfall not coinciding with the period for which lowest groundwater recharge was computed.
the second recharge method, periods of highest and lowest annual rainfall coincided with
periods of highest and lowest annual average recharge estimates. Year 2003 in which the
highest rainfall amount of the series was recorded coincided with the period of
annual average recharge estimate of 578.76mm. The same scenario was established for the
period of lowest annual rainfall and lowest annual average recharge estimate. The period with
Furthermore, most of the estimated annual recharge based on Sinha and Sharma method
year mean of 199.7 mm and 374.4 mm
Fig.2: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Sinha and
Fig.3: Annual and Long Term Mean of Recharge Estimate at Bauchi based on Turc Method
based on method 1 show that the period with
highest annual rainfall amount of the series (rainfall amount of 1416.1mm for 2003)
coincided with the period of the highest annual recharge estimate of 252.49mm. On the other
west annual rainfall of the series (615.8mm recorded in
2007) did not coincide with period of the lowest annual recharge estimate of 142.29mm
which was computed for 2001. The year 2001 for which the lowest annual recharge estimate
th the period with the second lowest annual rainfall of the series.
High potential evapotranspiration may a contributive factor to the period of lowest annual
rainfall not coinciding with the period for which lowest groundwater recharge was computed.
the second recharge method, periods of highest and lowest annual rainfall coincided with
periods of highest and lowest annual average recharge estimates. Year 2003 in which the
highest rainfall amount of the series was recorded coincided with the period of the highest
annual average recharge estimate of 578.76mm. The same scenario was established for the
period of lowest annual rainfall and lowest annual average recharge estimate. The period with
Journal of Geographic Issues, Vol. 2 No. 1, 2014
the lowest annual rainfall of the series (year 2007) coincide
annual average recharge estimate of 69.52mm.
Comparison of the estimated annual recharge with the long term mean based on Sinhna and
Sharma method shows that 7 out the 11 years had recharge estimates higher than the mean
value of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had
annual recharge estimates lower than and higher than the long term mean of 236.8mm.
Fig.4: Annual and Long Term Mean of Recharge Estimate at Gusau based on Sinha and
Sharma Method
Fig.5: Annual and Long Term Mean of Recharge Estimate at Gusau based on Turc Method
Variability and Equality of VarianceThe coefficient of variation for rainfall and recharge estimates for both locations is as
presented in Table 2.
Table 2: Coefficient of Variation of Rainfall and Recharge Estimates
Location Rainfall (mm)
Bauchi 21%
Gusau 22%
of Geographic Issues, Vol. 2 No. 1, 2014
65
the lowest annual rainfall of the series (year 2007) coincided with the period with the lowest
annual average recharge estimate of 69.52mm.
Comparison of the estimated annual recharge with the long term mean based on Sinhna and
Sharma method shows that 7 out the 11 years had recharge estimates higher than the mean
alue of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had
annual recharge estimates lower than and higher than the long term mean of 236.8mm.
Fig.4: Annual and Long Term Mean of Recharge Estimate at Gusau based on Sinha and
Fig.5: Annual and Long Term Mean of Recharge Estimate at Gusau based on Turc Method
Variability and Equality of Variance The coefficient of variation for rainfall and recharge estimates for both locations is as
: Coefficient of Variation of Rainfall and Recharge Estimates
Rainfall (mm) Sinha and Sharma (Method 1) Turc (Method 2)
17 % 47%
14% 56%
d with the period with the lowest
Comparison of the estimated annual recharge with the long term mean based on Sinhna and
Sharma method shows that 7 out the 11 years had recharge estimates higher than the mean
alue of 204.3mm (Fig.4). For Turc method (Fig.5) an almost equal number of years had
annual recharge estimates lower than and higher than the long term mean of 236.8mm.
Fig.4: Annual and Long Term Mean of Recharge Estimate at Gusau based on Sinha and
Fig.5: Annual and Long Term Mean of Recharge Estimate at Gusau based on Turc Method
The coefficient of variation for rainfall and recharge estimates for both locations is as
Turc (Method 2)
Journal of Geographic Issues, Vol. 2 No. 1, 2014
66
As shown in the Table the CV for both locations based on the Turc method is much higher
than coefficient of variation based on Sinha and Sharma method. This is an indication that
CV derived based on Turc method is more variable. Due to the inherent variability of climatic
and surface conditions, temporal variation of groundwater recharge occurs over different
temporal scales. These variations are normal and they depict the dynamic balance between
precipitation, infiltration, evaporation, transpiration, runoff, groundwater storage, withdrawal
and base-flow discharge (Resse and Risser, 2010). Apart from these factors, drought,
irrigation, agriculture and other forms of land use changes may also contribute to the
temporal variation of recharge within the study locations.
The comparison of the CV of rainfall with CV of recharge estimates for both methods shows
a closer relationship between rainfall and recharge estimate based on Sinha and Sharma
method.Results of the Levene’s test for equality of variance for Bauchi and Gusau are
represented Tables 3 and 4 respectively.
Table 3: Levene’s Test for Equality of Variance at Bauchi
Levene
Statistic df1 df2 Sig.
Recharge Based on Mean 11.462 1 20 .003
Based on Median 7.032 1 20 .015
Based on Median and
with adjusted df
7.032 1 10.859 .023
Based on trimmed
mean
10.716 1 20 .004
Table 4: Levene’s Test for Equality of Variance at Gusau
Levene
Statistic df1 df2 Sig.
Recharge Based on Mean 4.484 1 20 .047
Based on Median 4.145 1 20 .055
Based on Median and
with adjusted df
4.145 1 10.824 .067
Based on trimmed
mean
4.240 1 20 .053
For Bauchi, F(1, 20) = 11.46, P <0.05, where P = 0.003 and Gusau, F(1, 20) = 4.48, P <0.05
where P = 0.047, the assumption of equality or homogeneity of variance has been violated ,
indicating that the variance of the recharge estimates obtained from Sinha and Sharma
method, and those obtained from Turc method are significantly different from each other.
Regression Analysis The results of the regression model for Bauchi based on annual rainfall and recharge
estimates computed using the first empirical method is presented in Tables 5a to 5c, while the
regression model obtained from annual rainfall and recharge estimates from the second
empirical method is presented in Tables 6a to 6c.
Journal of Geographic Issues, Vol. 2 No. 1, 2014
67
As shown in Table 5a, R2 equals 0.984 indicating that rainfall accounts for 98.4 percent of
variation in natural groundwater recharge for the study period. Other factors that could be
responsible for the variability include vegetation and soil properties. The F-ratio (552.799) in
Table 5b is significant at p < 0.01 because the value in the column labeled sig (last column) is
less than 0.01. This indicates that there is less than 1 percent chance that an F-ratio this large
could be a chance occurrence, leading to the conclusion that the regression model results in
significantly better prediction.
The value of b0 (39.59) in Table 6c indicates that when there is no rainfall (when Xi= 0), the
model predicts that recharge will be 39.59mm while the value of b1 (Table 6c) which
represents the gradient of the regression line indicates that if there is an increase in rainfall by
1mm, then the regression model predicts that recharge will increase by 0.142mm.
The mean rainfall value of 1,130.41mm and the regression equation Ƴ = (b0 + b1X1) + ɛi was
used to predict the mean recharge for the 11-year study period, When:
Y = Mean recharge for the 11-year period
b0 = 39.59mm
bi = 0.142mm
Xi = 1, 130.41mm
is expressed as:
Mean recharge = (39.59 + 0.142 x 1, 130.41) = 200.11mm
Table 5a: Summary of Regression Model
Table 5b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance Value
ANOVAb
Model
Sum of
Squares df Mean Square F Sig.
1 Regression 11579.488 1 11579.488 552.799 .000a
Residual 188.523 9 20.947
Total 11768.011 10
a. Predictors: (Constant), Rainfall
b. Dependent Variable: recharge
Table 5c: Beta Values of Regression Model
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig. B Std. Error Beta
1 (Constant) 39.590 6.948 5.698 .000
Model Summary
Model R R Square
Adjusted R
Square
Std. Error of
the Estimate
1 .992a .984 .982 4.57679
a. Predictors: (Constant), Rainfall
Journal of Geographic Issues, Vol. 2 No. 1, 2014
68
Rainfall .142 .006 .992 23.512 .000
a. Dependent Variable: recharge
With regards to the outcome of the regression model of annual rainfall and recharge estimates
obtained from the second empirical method, Table 6a shows that rainfall accounts for 99
percent (R2= 0.996) of the variation in groundwater recharge. Table 6b indicates that the F-
ratio is significant at P < 0.01, while b0 predicts that when there is no rainfall (when Xi = 0),
the recharge value will be -453.70mm. Such negative value is however impossible in nature.
This is because during periods of rainfall cessation such as the dry season, natural
groundwater recharge can occur from seepage from surface water bodies such as streams,
lakes or ponds, irrigation return flow, inter-aquifer flows and urban recharge such as leaky
water and sewage systems. bi which is the slope of the regression line indicates that there will
be an increase of 0.733mm in recharge for every 1mm increase in rainfall.
The prediction of the 11-year mean recharge based on a mean rainfall value of 1,130.41mm
(ie Xi = 1, 130.41mm) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:
Y = Mean recharge for the 11-year period
b0 = -453.70mm
bi = 0.733mm
is expressed as:
Mean recharge = (-453.70 + 0.733 x 1, 130.41) = 374.89mm
Table 6a: Summary of Regression Model
Model R R Square
Adjusted R
Square
Std. Error of
the Estimate
1 .998a .996 .996 11.80280
a. Predictors: (Constant), Rainfall
Table 6b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance
Value
ANOVAb
Model
Sum of
Squares df Mean Square F Sig.
1 Regression 309790.897 1 309790.897 2223.815 .000a
Residual 1253.755 9 139.306
Total 311044.652 10
a. Predictors: (Constant), Rainfall
b. Dependent Variable: Recharge
Table 6c: Beta Values of Regression Model
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig. B Std. Error Beta
Journal of Geographic Issues, Vol. 2 No. 1, 2014
69
1 (Constant) -453.700 17.917 -25.322 .000
Rainfall .733 .016 .998 47.157 .000
a. Dependent Variable: Recharge
The results of the regression model for Gusau based on annual rainfall and recharge estimates
computed using the first empirical method is presented in Tables 7a to 7c, while the
regression model obtained from annual rainfall and recharge estimates computed with the
second empirical method is presented in Tables 8a to 8c.
As shown in Table 7a, the value of R2 (0.616) indicates that rainfall accounts for 61.6 percent
of the variation in natural groundwater recharge. The F-ratio (14.447) as shown in Table 7b is
significant at P < 0.01 because the value 0.004 in the column labeled sig is less than 0.01.
The value b0 (102.612) in Table 8c indicates that when there is no rainfall (when Xi = 0), the
model predicts a recharge of 102.62mm.The slope of the regression line (bi) in Table 7c
indicates that when there an increase of 1mm in rainfall, groundwater recharge will increase
by 0.11mm.
The prediction of the 11-year mean recharge based on a mean rainfall value of 922.70mm (ie
Xi = 922.70) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:
Y = Mean recharge for the 11-year period
b0 = 102.62mm
bi = 0.11mm
is expressed as:
Mean recharge = (102.62 + 0.11 x 922.70) = 204.12mm
Table 7a: Summary of Regression Model
Table 7b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance Value
ANOVAb
Model Sum of Squares df Mean Square F Sig.
1 Regression 5207.472 1 5207.472 14.447 .004a
Residual 3244.051 9 360.450
Total 8451.523 10
a. Predictors: (Constant), Rainfall
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig. B Std. Error Beta
1 (Constant) 102.619 27.359 3.751 .005
Model Summary
Model R R Square
Adjusted R
Square
Std. Error of
the Estimate
1 .785a .616 .574 18.98552
a. Predictors: (Constant), Rainfall
Table 7c: Beta Values of Regression Model
Journal of Geographic Issues, Vol. 2 No. 1, 2014
70
Rainfall .110 .029 .785 3.801 .004
a. Dependent Variable: Recharge
For the regression model of annual rainfall and recharge estimates obtained from the second
empirical method, Table 8a indicates that rainfall accounts for 98 percent (R2 = 0.980) of the
variation in natural groundwater recharge for the study period. The F-ratio (448.948) in Table
8b is significant at P < 0.01 because the value 0.000 under the column labeled sig is lesser
0.01, while the value of b0 predicts that when there is no rainfall (when Xi = 0), the recharge
value will be -350.60mm. The negative value however, does not mean the cessation of
recharge for reasons earlier explained. The value of bi predicts that for every 1mm increase in
rainfall, there will be a corresponding increase in recharge by 0.637mm.
The prediction of the 11-year mean recharge based on a mean rainfall value of 922.70mm (ie
Xi = 922.70) and the regression equation Ƴ = (b0 + b1X1) + ɛi , when:
Y = Mean recharge for the 11-year period
b0 = -350.597mm
bi = 0.637mm
is expressed as:
Mean recharge = (-350.59 + 0.637 x 922.70) = 237.16mm
Table 8a: Summary of Regression Model
Table 8b: Analysis of Variance, Sums of Squares, F-Ratio and Associated Significance
Value
Table 8c: Beta Values of Regression Model
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig. B Std. Error Beta
1 (Constant) -350.597 28.349 -12.367 .000
Rainfall .637 .030 .990 21.188 .000
Model Summary
Model R R Square
Adjusted R
Square
Std. Error of
the Estimate
1 .990a .980 .978 19.67218
a. Predictors: (Constant), Rainfall
ANOVAb
Model
Sum of
Squares df Mean Square F Sig.
1 Regression 173740.586 1 173740.586 448.948 .000a
Residual 3482.953 9 386.995
Total 177223.540 10
a. Predictors: (Constant), Rainfall
b. Dependent Variable: Recharge
Journal of Geographic Issues, Vol. 2 No. 1, 2014
71
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig. B Std. Error Beta
1 (Constant) -350.597 28.349 -12.367 .000
Rainfall .637 .030 .990 21.188 .000
a. Dependent Variable: Recharge
CONCLUSION AND RECOMMENDATION The study has shown that empirical methods can be used as a plausible means of estimating
groundwater recharge. The moderate (Sinha and Sharma Method) to high level (Turc
Method) of variability as shown by the coefficient of variation is an indication that the
variable nature of recharge should be given more consideration in the development of
groundwater. This would help in preventing groundwater mining and ensure sustainable
water use.
The study also showed through regression analysis that rainfall generally accounts for more
than 90 percent of groundwater variation in the study locations. This therefore shows that
statistical methods such as regression analysis can serve as a veritable means of predicting the
rate of groundwater recharge.
Going by the results of this study, it is recommended that scientific and periodic assessments
of groundwater resources potential should be embarked upon. Furthermore, more efforts
should be aimed at water conservation through rain water harvesting in the wet season to
reduce the level of reliance on groundwater.
REFERENCES
Adelana, S.M.A.., Olasehinde, P.I. and Vrbka, P. (2006). A quantitative estimation of
groundwater
recharge in part of the Sokoto Basin, Nigeria. Journal of Environmental Hydrology,
14(5), 1-16.
Adeleke, O.O., Makinde, V., Eruola, A.O., Dada, O.F., Akintayo, O.O., and Aluko, T.J.
(2015).
Estimation of groundwater recharges using empirical formulae in Odeda Local
Government Area, Ogun State Nigeria. Challenges, 6, 271-281
Dalhatu, S., and Garba, J. (2012). Soil resources degradation and conservation techniques
adopted
among the small holder farmers in Gusau, North-western Nigeria. Nigerian Journal of
Basic and Applied Sciences, 20 (3), 134-141.
De Vries., J.J. and Simmers, I. (2002). Groundwater recharge: An overview of process and
challenges. Hydrogeology Journal, 10, 5-17.
Dike, E.F.C., Aga, T. and Bashayi, T.E.O. (2008). Geochemical investigation of groundwater
in
shallow hand dug wells in wells in railway area, Bauchi. Continental Journal of
Applied Sciences, 3, 57-64.
Journal of Geographic Issues, Vol. 2 No. 1, 2014
72
Du Preez, J.W., and Barber, W. (1965). The distribution and chemical quality of
groundwater in
Northern Nigeria. (Lagos: Federal Government of Nigeria).
Edmunds, W.M.., Fellman, E. Goni, L.B. and Prudhomme, C. (2002). Spatial and temporal
distribution of groundwater recharge in northern Nigeria. Hydrogeology Journal, 10,
205-215.
Goes, B.J.M., (1999). Estimate of shallow groundwater recharge in the Hadejia-Nguru
wetlands,
Semi-arid, Northeast, Nigeria. Hydrogeology Journal, 7, 305-316.
Hazell, J.R.T., (1985). Development of groundwater for farming communities in Bauchi
State,
Nigeria. In hydrogeology in the service of man. Memoirs of the 18th
congress of the
International Association of Hydrogeologists, 8-13 September, 1985, Cambridge:
IAHS Publication No 154.
Hill, I.D. (Ed). (1978). Land resources of central Nigeria: Agricultural development
possibilities. Vol 1B. The Bauchi Plains. Surrey: Land Resources Development
Centre.
Kumar, C.P. (2000). Groundwater assessment methodology. National Institute of Hydrology.
Kumar, C.P., and Seethapathi, P.V. (2002). Assessment of natural groundwater recharge in
upper
Ganga command area. Journal of Applied Hydrology, 14(4), 13-20.
Ladan, M.U., Abubakar, U. Abdullahi, K. Bunza. M.D.A. Ladan, M.J. and Adamu, T.
(2012).
Studies on urinary schistosomiasis in selected villages around Gusau Dam site,
Zamfara State. Nigeria. Nigerian Journal of Basic and Applied Sciences, 20(3), 189-
194.
Oke, M.O., Martins, O., Idowu, O., and Aiyelokun, O. (2013). Comparative analysis of
empirical
formulae used in groundwater recharge in Ogun-Oshun River Basin. Journal of
Scientific Research and Reports, 2(2), 692-710.
Rawat, K.S., Mishra, A.K., Paul, G., Kumar, R. (2012). Estimation of groundwater recharge
in Shankergarh block of Allahabad, India using remote sensing and statistical approach.
Global Journal of Science, Engineering and Technology, 1, 34-48.
Sharp, J.M., (2007). The glossary of hydrogeological terms. The University of Texas, Texas.
Usman, H.A., and Mohammed, B.U. (2012). Solid waste management in Bauchi, Nigeria
(Obstacles and prospects). Journal of Environmental Science and Resources
Management, 4, 96-107.